A power battery safety detection and evaluation system based on multi-modal visualization

Abstract

The application relates to the technical field of battery safety detection, and discloses a power battery safety detection and evaluation system based on multi-modal visualization, which comprises a detection device and an evaluation device, wherein the detection device is connected with the evaluation device; the detection device comprises a sealed test cavity, a safety trigger and loading module, a multi-modal parameter acquisition module and a data processing and visualization module; an analysis module constructs a coupling characteristic vector based on temperature characteristics, smoke characteristics and deformation characteristics, and determines a safety evaluation value of a battery to be detected according to the coupling characteristic vector; a judgment optimization module judges whether the safety evaluation value needs to be optimized according to power data; if yes, an optimization coefficient of the safety evaluation value is determined based on the air pressure in the bin and the power data, and an optimized safety evaluation value is obtained; and a judgment module determines the safety grade of the battery to be detected according to the optimized safety evaluation value. The application can comprehensively, accurately and intuitively detect and evaluate the safety condition of the power battery.

Description

Technical Field

[0001] The present invention relates to the technical field of battery safety detection, and more particularly, to a power battery safety detection and evaluation system based on multi-modal visualization. Background Art

[0002] With the rapid increase in the penetration rate of new energy vehicles, the power battery safety issue has shifted from "whether it is safe" to "how to visualize and quantitatively evaluate safety". Most of the existing laboratory testing devices focus on single working conditions or parameters, and the acquisition and display links are fragmented, making it difficult to restore the true evolution of the coupled processes of heat, electricity, force, and gas. The data application uses simple thresholds or linear weighted scoring, which is difficult to be migrated to different battery systems and complex scenarios, resulting in insufficient interpretability, comparability, and reproducibility of the results.

[0003] At the same time, the battery object has expanded from the traditional lithium-ion system to all-solid-state batteries. When the solid-state system is abnormal, it is accompanied by corrosive or toxic gases and high-temperature solid / gas two-phase coupling behaviors, which pose higher requirements for the materials of the test chamber. Most of the existing devices are general metal chamber and conventional gas sensor solutions, which are difficult to take into account properties such as temperature resistance and corrosion resistance. Moreover, the existing system only stays at the simple monitoring level for "visualization" and cannot perform linkage analysis of the thermal field - flue gas - mechanical deformation, affecting the judgment of battery safety.

[0004] At the test execution level, there are many problems with traditional platforms: the chamber volume and the temperature resistance / pressure resistance margin are limited, making it difficult to cover the heat release and gas release peaks of large-capacity single cells; the efficiency and consistency of the opening / closing and sealing links are insufficient; the trigger / loading units are scattered, and the parameter resolution is limited, making it difficult to complete the programmable working condition path scanning; there is no unified model for data acquisition, resulting in high costs for subsequent analysis and comparison. These factors restrict the systematic understanding of battery safety, especially the cross-system and cross-scenario consistency evaluation.

[0005] Therefore, it is necessary to design a power battery safety detection and evaluation system based on multi-modal visualization to solve the problems existing in the current technology. Summary of the Invention

[0006] In view of this, the present invention proposes a power battery safety detection and evaluation system based on multi-modal visualization, aiming to solve the above-mentioned existing problems.

[0007] The present invention proposes a power battery safety detection and evaluation system based on multi-modal visualization, comprising:

[0008] A detection device and an evaluation device, the detection device is connected to the evaluation device; the detection device includes a sealed test chamber, a safety trigger and loading module, a multi-modal parameter acquisition module, and a data processing and visualization module; the evaluation device includes an analysis module, a judgment and optimization module, and a determination module; wherein,

[0009] The battery to be tested is placed inside the sealed test chamber;

[0010] The safety triggering and loading module is used to apply trigger loading to the battery under test under controlled conditions and obtain the response data of the battery under test under the trigger loading conditions;

[0011] The multimodal parameter acquisition module is used to acquire response data and perform preprocessing.

[0012] The data processing and visualization module is used to draw superimposed thermal field maps, flue gas isosurfaces, and deformation curves of the battery under test based on the preprocessed response data.

[0013] The analysis module is used to extract the temperature features of the superimposed thermal field map, the flue gas features of the flue gas isosurface, and the deformation features of the deformation curve. Based on the temperature features, flue gas features, and deformation features, a coupled feature vector is constructed, and the safety assessment value of the battery to be tested is determined according to the coupled feature vector.

[0014] The optimization module is used to collect the power data of the battery under test and determine whether the safety assessment value should be optimized based on the power data. If so, it collects the air pressure inside the sealed test chamber, determines the optimization coefficient of the safety assessment value based on the air pressure inside the chamber and the power data, and obtains the optimized safety assessment value.

[0015] The determination module is used to determine the safety level of the battery to be tested based on the optimized safety assessment value.

[0016] Furthermore, the controlled conditions include any one or more combinations of overcharging / over-discharging, heating, squeezing, and puncture triggering of the battery under test.

[0017] Furthermore, when plotting the superimposed thermal field map, flue gas isosurface, and deformation curve of the battery under test based on the preprocessed response data, the following steps are included:

[0018] The response data includes temperature data, flue gas concentration data, and deformation data;

[0019] The preprocessed temperature data is processed, and thermal imaging algorithms are used to map temperature values ​​at different times and locations onto a two-dimensional plane. The temperature is represented by the intensity of color to create an overlayable thermal field map.

[0020] The preprocessed flue gas concentration data is processed, and the isosurface generation algorithm is used to determine the spatial distribution boundary of the flue gas and draw the flue gas isosurface.

[0021] The preprocessed deformation data is processed, and a deformation curve is plotted by fitting a curve with time as the horizontal axis and deformation as the vertical axis.

[0022] Furthermore, a coupled feature vector is constructed based on temperature characteristics, flue gas characteristics, and deformation characteristics. When determining the safety assessment value of the battery under test based on the coupled feature vector, the following steps are included:

[0023] Temperature characteristics, flue gas characteristics, and deformation characteristics are normalized separately to obtain normalized characteristics;

[0024] Feature enhancement is performed on the normalized features to obtain enhanced features;

[0025] The enhanced features are aligned temporally and spatially to form an aligned multimodal feature set;

[0026] Based on the multimodal feature set, construct the coupling terms between modes, and combine the coupling terms with the multimodal feature set to form the coupling feature vector;

[0027] The coupling feature vector is compared with the historical coupling feature vector set, and the safety assessment value of the battery to be tested is determined based on the comparison result.

[0028] The coupling terms include the coupling terms of temperature change rate and flue gas diffusion gradient, the coupling terms of peak flue gas concentration and deformation accumulation rate, and the coupling terms of temperature spatial abrupt change and deformation inflection point time sequence difference.

[0029] Furthermore, when determining the safety assessment value of the battery to be tested based on the comparison results, the following are included:

[0030] When there is a historical coupling feature vector in the historical coupling feature vector group that is the same as the coupling feature vector, the historical security assessment value corresponding to the historical coupling feature vector is used as the security assessment value.

[0031] When there is no historical coupling feature vector in the historical coupling feature vector group that is the same as the coupling feature vector, the state offset rate of the battery to be tested is calculated based on the coupling feature vector, and the safety assessment value of the battery to be tested is determined based on the state offset rate.

[0032] Furthermore, when calculating the state shift rate of the battery to be detected based on the coupling feature vector, the following steps are included:

[0033] Obtain the time series of the coupled feature vectors, and calculate the Euclidean distance between the coupled feature vectors at adjacent time points based on the time series of the coupled feature vectors to obtain the amount of coupling change;

[0034] The coupling changes are weighted, with the weights consisting of temperature characteristic weights, flue gas characteristic weights, and deformation characteristic weights, to obtain the weighted coupling changes.

[0035] The rate of change of the weighted coupling change is calculated within a preset time window to obtain the state offset rate.

[0036] Furthermore, when determining the safety assessment value of the battery under test based on the state offset rate, the following steps are included:

[0037] The state offset rate is compared with the first state offset rate and the second state offset rate, and the safety assessment value is determined based on the comparison result; wherein the first state offset rate is less than the second state offset rate.

[0038] When the state offset rate is less than or equal to the first state offset rate, the security assessment value is determined to be the first security assessment value.

[0039] When the state offset rate is greater than the first state offset rate and less than or equal to the second state offset rate, the security assessment value is determined to be the second security assessment value.

[0040] When the state offset rate is greater than the second state offset rate, the security assessment value is determined to be the third security assessment value.

[0041] Furthermore, when determining whether to optimize the safety assessment value based on power data, the following are included:

[0042] The power data is analyzed, and noise reduction and filtering are performed on the power data;

[0043] Power mutation indicators are obtained based on filtered power data. These indicators include amplitude mutation rate, relative change rate, and duration.

[0044] The power mutation index is compared with a preset threshold. Only when the power mutation index meets the preset amplitude threshold and exceeds a preset number of times within a preset time window is it determined that there is a power mutation value in the power data.

[0045] When a sudden power fluctuation is detected in the power data, it is determined that the safety assessment value needs to be optimized.

[0046] When it is determined that there are no power mutation values ​​in the power data, it is determined that there is no need to optimize the safety assessment value.

[0047] Furthermore, when determining the optimization coefficients for the safety assessment value based on the air pressure and electricity data inside the warehouse, and obtaining the optimized safety assessment value, the following steps are taken:

[0048] The air pressure inside the chamber was analyzed to obtain the rate of change of air pressure.

[0049] The power data is analyzed to obtain real-time current values;

[0050] The pressure change rate and pressure change threshold are compared, and the real-time current value is compared with the current threshold. The optimization coefficient is determined based on the comparison results.

[0051] When the rate of change of air pressure is greater than or equal to the air pressure change threshold, and the real-time current value is greater than or equal to the current threshold, the optimization coefficient is determined as the first optimization coefficient.

[0052] When the rate of change of air pressure is greater than or equal to the air pressure change threshold and the real-time current value is less than the current threshold, the optimization coefficient is determined as the second optimization coefficient.

[0053] When the rate of change of air pressure is less than the air pressure change threshold and the real-time current value is greater than or equal to the current threshold, the optimization coefficient is determined as the third optimization coefficient.

[0054] When the rate of change of air pressure is less than the air pressure change threshold and the real-time current value is less than the current threshold, the optimization coefficient is determined to be the fourth optimization coefficient.

[0055] The product of the optimization coefficient and the safety assessment value is used as the optimized safety assessment value.

[0056] Furthermore, when determining the safety level of the battery under test based on the optimized safety assessment value, the following are included:

[0057] The optimized safety assessment value is compared with the preset safety level mapping table, and the safety level of the battery to be tested is determined based on the comparison results.

[0058] Compared with existing technologies, the beneficial effects of this invention are as follows: The multimodal visualization-based power battery safety detection and evaluation system provided by this invention can comprehensively, accurately, and intuitively detect and evaluate the safety status of power batteries. Regarding the detection device, the sealed test chamber provides a stable and controllable environment, avoiding external interference; the safety triggering and loading module simulates extreme battery conditions and acquires response data; the multimodal parameter acquisition module collects and preprocesses data to ensure accuracy and usability; the data processing and visualization module analyzes the data and draws overlayable thermal field diagrams, flue gas isosurfaces, and deformation curves for convenient observation and analysis. Regarding the evaluation device, the analysis module extracts graphic features to construct coupled feature vectors, determines safety evaluation values, and considers multiple features to make the results more scientific; the judgment and optimization module optimizes the evaluation values ​​based on power data and chamber air pressure to improve accuracy; the judgment module determines the battery safety level based on the optimized evaluation values, providing a reference basis. Compared with existing technologies, this system's multimodal visualization method is more comprehensive, with each module collaborating to form a complete system, ensuring reliable and accurate detection results. This is of great significance for improving power battery safety and ensuring stable operation, and can play a role in multiple fields, reducing risk losses. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 This is a structural block diagram of a power battery safety testing and evaluation system based on multimodal visualization, provided in an embodiment of the present invention. Detailed Implementation

[0061] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0062] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0063] See Figure 1 As shown in some embodiments of this application, this embodiment provides a power battery safety testing and evaluation system based on multimodal visualization, including:

[0064] The system includes a testing device and an evaluation device, with the testing device connected to the evaluation device. The testing device comprises a sealed test chamber, a safety triggering and loading module, a multimodal parameter acquisition module, and a data processing and visualization module. The evaluation device includes an analysis module, a judgment and optimization module, and a decision module.

[0065] The battery to be tested is placed inside the sealed test chamber;

[0066] The safety triggering and loading module is used to apply trigger loading to the battery under test under controlled conditions and obtain the response data of the battery under test under the trigger loading conditions;

[0067] The multimodal parameter acquisition module is used to acquire response data and perform preprocessing.

[0068] The data processing and visualization module is used to draw superimposed thermal field maps, flue gas isosurfaces, and deformation curves of the battery under test based on the preprocessed response data.

[0069] The analysis module is used to extract the temperature features of the superimposed thermal field map, the flue gas features of the flue gas isosurface, and the deformation features of the deformation curve. Based on the temperature features, flue gas features, and deformation features, a coupled feature vector is constructed, and the safety assessment value of the battery to be tested is determined according to the coupled feature vector.

[0070] The optimization module is used to collect the power data of the battery under test and determine whether the safety assessment value should be optimized based on the power data. If so, it collects the air pressure inside the sealed test chamber, determines the optimization coefficient of the safety assessment value based on the air pressure inside the chamber and the power data, and obtains the optimized safety assessment value.

[0071] The determination module is used to determine the safety level of the battery to be tested based on the optimized safety assessment value.

[0072] It is understood that the multimodal visualization-based power battery safety detection and evaluation system provided in this embodiment can comprehensively, accurately, and intuitively detect and evaluate the safety status of power batteries. Regarding the detection device, the sealed test chamber provides a stable and controllable environment, avoiding external interference; the safety triggering and loading module simulates extreme battery conditions and acquires response data; the multimodal parameter acquisition module collects and preprocesses data to ensure accuracy and usability; the data processing and visualization module analyzes the data and draws overlayable thermal field diagrams, flue gas isosurfaces, and deformation curves for easy observation and analysis. Regarding the evaluation device, the analysis module extracts graphic features to construct coupled feature vectors, determines the safety evaluation value, and considers multiple features to make the results more scientific; the judgment and optimization module optimizes the evaluation value based on power data and chamber air pressure to improve accuracy; the judgment module determines the battery safety level based on the optimized evaluation value, providing a reference basis. Compared with existing technologies, this system's multimodal visualization method is more comprehensive, with each module collaborating to form a complete system, ensuring reliable and accurate detection results. This is of great significance for improving power battery safety and ensuring stable operation, and can play a role in multiple fields, reducing risk and loss.

[0073] Specifically, the controlled conditions include any one or more combinations of overcharging / over-discharging, heating, squeezing, and puncture of the battery under test.

[0074] Understandably, these controlled conditions are set to simulate as many extreme situations the battery under test might encounter in actual use as possible, thereby providing a more comprehensive assessment of the battery's safety performance. Overcharging / over-discharging can disrupt the chemical balance within the battery, potentially leading to overheating, bulging, or even explosion; heating can trigger thermal runaway, causing serious safety incidents; and compression and puncture can damage the battery's internal structure, causing short circuits and other problems. By combining these triggering conditions, the battery's safety status under complex environments can be more realistically reflected. Furthermore, during these triggering operations, the safety triggering and loading module precisely controls the intensity and timing of the triggering to ensure accurate response data. The entire process is conducted within a sealed test chamber, which not only prevents potential leakage of hazardous substances but also ensures the stability and consistency of the test environment. In addition, for different types and specifications of batteries under test, the combination and parameters of the controlled conditions can be flexibly adjusted according to their characteristics and application scenarios to achieve more precise safety testing and evaluation.

[0075] Specifically, when plotting the superimposed thermal field map, flue gas isosurface, and deformation curve of the battery under test based on the preprocessed response data, the following steps are included:

[0076] The response data includes temperature data, flue gas concentration data, and deformation data;

[0077] The preprocessed temperature data is processed, and thermal imaging algorithms are used to map temperature values ​​at different times and locations onto a two-dimensional plane. The temperature is represented by the intensity of color to create an overlayable thermal field map.

[0078] The preprocessed flue gas concentration data is processed, and the isosurface generation algorithm is used to determine the spatial distribution boundary of the flue gas and draw the flue gas isosurface.

[0079] The preprocessed deformation data is processed, and a deformation curve is plotted by fitting a curve with time as the horizontal axis and deformation as the vertical axis.

[0080] Understandably, this method of plotting overlayable thermal field maps, flue gas isosurfaces, and deformation curves can visually represent the previously abstract data on temperature, flue gas concentration, and deformation in an intuitive graphical form. Overlayable thermal field maps allow operators to clearly see the temperature changes of the battery at different times and locations, and through color contrasts, quickly identify high-temperature areas of the battery, which is crucial for determining whether the battery is at risk of overheating. Flue gas isosurfaces visually present the spatial distribution of flue gas, helping to understand the diffusion of flue gas generated by internal chemical reactions within the battery, providing a basis for analyzing the intensity of internal reactions and potential safety hazards. Deformation curves, using time and deformation as coordinate axes, clearly show the deformation trend of the battery during the trigger loading process. Operators can judge the structural stability of the battery based on the curve's trend and predict whether the battery will rupture or cause serious safety accidents.

[0081] Specifically, when constructing a coupled feature vector based on temperature characteristics, flue gas characteristics, and deformation characteristics, and determining the safety assessment value of the battery under test based on the coupled feature vector, the process includes:

[0082] Temperature characteristics, flue gas characteristics, and deformation characteristics are normalized separately to obtain normalized characteristics;

[0083] Feature enhancement is performed on the normalized features to obtain enhanced features;

[0084] The enhanced features are aligned temporally and spatially to form an aligned multimodal feature set;

[0085] Based on the multimodal feature set, construct the coupling terms between modes, and combine the coupling terms with the multimodal feature set to form the coupling feature vector;

[0086] The coupling feature vector is compared with the historical coupling feature vector set, and the safety assessment value of the battery to be tested is determined based on the comparison result.

[0087] The coupling terms include the coupling terms of temperature change rate and flue gas diffusion gradient, the coupling terms of flue gas concentration peak and deformation accumulation rate, and the coupling terms of temperature spatial abrupt change and deformation inflection point time sequence difference.

[0088] Understandably, this method of constructing coupled feature vectors and determining safety assessment values ​​fully considers the interrelationships among the three different modal features: temperature, flue gas, and deformation. Normalization eliminates dimensional differences between features, enabling comparison and analysis on the same scale. Feature enhancement further highlights important information and improves feature discriminability. Temporal and spatial alignment ensures consistency between different modal features in time and space, making the constructed multimodal feature set more reasonable and accurate.

[0089] Understandably, the coupling term between the rate of temperature change and the flue gas diffusion gradient reflects the correlation between battery thermal effects and flue gas diffusion. For example, if the battery temperature rises rapidly during trigger loading, with a large rate of temperature change, and the flue gas diffusion gradient is also large at this time, it means that the temperature rise causes the flue gas generated by the chemical reaction within the battery to diffuse rapidly. This suggests a violent reaction, abundant gaseous products, and accelerated heat release leading to flue gas diffusion. Continued operation in this situation may indicate a risk of thermal runaway, reducing the safety assessment value. The coupling term between the peak flue gas concentration and the deformation accumulation rate reflects the relationship between the amount of flue gas generated and the degree of battery deformation. In a certain test, if the flue gas concentration reaches its peak and the deformation accumulation rate is large, it indicates that a large amount of flue gas is generated within the battery and the structure is significantly deformed. This may indicate a violent chemical reaction and severe damage to the internal structure, potentially causing safety issues and affecting the safety assessment value. The coupling term between the spatial abruptness of temperature and the timing difference of the deformation inflection point focuses on the temporal relationship between spatial abruptness of temperature and the critical nodes of battery deformation. If the temperature in a localized area of ​​the battery rises sharply in a short period of time, and the time interval between this rise and the deformation inflection point is short, it indicates that the temperature change quickly triggers changes in the battery structure, which can easily lead to safety hazards and reduce the safety assessment value. Analyzing these coupling terms allows for a more comprehensive and accurate determination of the safety assessment value of the battery under test.

[0090] Understandably, for the coupling term between the rate of temperature change and the flue gas diffusion gradient, a weighted product method can be used to determine the coupling value. First, both the rate of temperature change and the flue gas diffusion gradient are standardized to a range of [0,1]. Then, different weights are assigned to them based on the actual situation; for example, the rate of temperature change is weighted at 0.6, and the flue gas diffusion gradient at 0.4. The standardized rate of temperature change and the flue gas diffusion gradient are multiplied by their respective weights, and the result is the coupling value of the coupling term. For the coupling term between the peak flue gas concentration and the cumulative deformation rate, a nonlinear function can be constructed to calculate the coupling value. First, the peak flue gas concentration and the cumulative deformation rate are normalized to a range of [0,1]. Then, a nonlinear function, such as a quadratic or exponential function, is fitted based on a large amount of experimental data. Substituting the normalized peak flue gas concentration and the cumulative deformation rate into this function yields the coupling value of the coupling term. For the coupling term between the spatial abrupt change in temperature and the time difference of the deformation inflection point, fuzzy logic can be used to determine the coupling value. Based on the degree of temperature spatial abruptness and the magnitude of the time difference in deformation inflection points, they are classified into different fuzzy levels, such as "high," "medium," and "low." Then, a fuzzy rule base is established; for example, when the temperature spatial abruptness is "high" and the time difference in deformation inflection points is "low," the coupling value corresponds to the value associated with "high risk." Through fuzzy inference, based on the actual temperature spatial abruptness and the level of the time difference in deformation inflection points, the coupling value of this coupling term is derived from the fuzzy rule base.

[0091] Specifically, when determining the safety assessment value of the battery to be tested based on the comparison results, the following are included:

[0092] When there is a historical coupling feature vector in the historical coupling feature vector group that is the same as the coupling feature vector, the historical security assessment value corresponding to the historical coupling feature vector is used as the security assessment value.

[0093] When there is no historical coupling feature vector in the historical coupling feature vector group that is the same as the coupling feature vector, the state offset rate of the battery to be tested is calculated based on the coupling feature vector, and the safety assessment value of the battery to be tested is determined based on the state offset rate.

[0094] Understandably, this method of determining safety assessment values ​​based on comparison results considers both the reference value of historical data and allows for flexible handling of new situations. When identical historical coupling feature vectors exist, directly using the corresponding historical safety assessment value yields quick and efficient results because historical data has already verified the correlation between the feature vector and the safety assessment value. However, when identical historical coupling feature vectors do not exist, determining the safety assessment value by calculating the state shift rate fully considers the differences between the current and historical states of the battery. The state shift rate reflects the drastic nature of battery state changes. A large state shift rate indicates a significant difference between the current state and historical conditions, potentially leading to higher safety risks, and consequently, a lower safety assessment value. Conversely, a small state shift rate indicates a relatively stable battery state, and a potentially higher safety assessment value.

[0095] Specifically, when calculating the state shift rate of the battery to be detected based on the coupling feature vector, the following steps are included:

[0096] Obtain the time series of the coupled feature vectors, and calculate the Euclidean distance between the coupled feature vectors at adjacent time points based on the time series of the coupled feature vectors to obtain the coupling change.

[0097] The coupling changes are weighted, with the weights consisting of temperature characteristic weights, flue gas characteristic weights, and deformation characteristic weights, to obtain the weighted coupling changes.

[0098] The rate of change of the weighted coupling change is calculated within a preset time window to obtain the state offset rate.

[0099] Understandably, suppose we have a battery to be tested, and we perform multiple tests on it over a period of time, obtaining a series of coupled feature vectors. Let the time series be t1, t2, t3...tn, and the corresponding coupled feature vectors be C1, C2, C3...Cn, respectively. First, calculate the Euclidean distance between the coupled feature vectors at adjacent time points. Taking time points t1 and t2 as an example, let C1 = (x1, y1, z1) and C2 = (x2, y2, z2), then the Euclidean distance between them is d12 = {[(x2-x1)...tn}. 2+(y2-y1) 2 +(z2-z1) 2 ]}1 / 2, this is the coupling change from time t1 to t2. Using the same method, calculate the coupling changes d23, d34, ..., dn-1n from time t2 to t3, t3 to t4, ..., tn-1 to tn sequentially. Next, perform weighted processing. Assume the weight of the temperature feature is 0.4, the weight of the flue gas feature is 0.3, and the weight of the deformation feature is 0.3. For each coupling change, weight is applied according to its corresponding feature. For example, for d12, if the contribution of the temperature feature change to the Euclidean distance is a, the contribution of the flue gas feature change is b, and the contribution of the deformation feature change is c, then the weighted coupling change Dw12 = 0.4a + 0.3b + 0.3c. Similarly, perform this weighted processing on all coupling changes to obtain Dw23, Dw34, ..., Dwn-1n. Finally, calculate the rate of change within a preset time window. Assuming the preset time window is from t1 to t3, the initial value of the weighted coupling change is Dw12, and the final value is Dw23 within this time window. The state shift rate S = (Dw23 - Dw12) / (t3 - t1). Through this calculation, we can obtain the state shift rate of the battery under test within the preset time window, and then determine the battery's safety assessment value based on this rate. This calculation method comprehensively considers the changes in the coupling feature vector over time, as well as the weights of different features, which can more accurately reflect the changes in the battery state and provide a more reliable basis for battery safety assessment.

[0100] Specifically, when determining the safety assessment value of the battery under test based on the state offset rate, the following are included:

[0101] The state offset rate is compared with the first state offset rate and the second state offset rate, and the safety assessment value is determined based on the comparison result; wherein the first state offset rate is less than the second state offset rate.

[0102] When the state offset rate is less than or equal to the first state offset rate, the security assessment value is determined to be the first security assessment value.

[0103] When the state offset rate is greater than the first state offset rate and less than or equal to the second state offset rate, the security assessment value is determined to be the second security assessment value.

[0104] When the state offset rate is greater than the second state offset rate, the security assessment value is determined to be the third security assessment value.

[0105] Understandably, the safety assessment values ​​are ranked as follows: First safety assessment value > Second safety assessment value > Third safety assessment value. When the state offset rate is less than or equal to the first state offset rate, it indicates that the battery state changes slowly and stably, and the battery faces a low safety risk; therefore, a higher first safety assessment value is given. When the state offset rate is between the first and second state offset rates, it indicates that the battery state changes more rapidly, but is still within an acceptable range. At this point, the battery faces a certain safety risk, so the safety assessment value is correspondingly reduced to the second safety assessment value. When the state offset rate is greater than the second state offset rate, it means that the battery state changes drastically, and there may have been a serious safety hazard, with the possibility of a safety accident at any time; therefore, the lowest third safety assessment value is given. This method of determining safety assessment values ​​based on comparing the state offset rate with different thresholds provides a quantitative and intuitive evaluation of the battery's safety status, helping relevant personnel to take timely and appropriate measures to ensure the safe use of the battery.

[0106] Specifically, when determining whether to optimize safety assessment values ​​based on power data, this includes:

[0107] The power data is analyzed, and noise reduction and filtering are performed on the power data;

[0108] Power mutation indicators are obtained based on filtered power data. These indicators include amplitude mutation rate, relative change rate, and duration.

[0109] The power mutation index is compared with a preset threshold. Only when the power mutation index meets the preset amplitude threshold and exceeds a preset number of times within a preset time window is it determined that there is a power mutation value in the power data.

[0110] When a sudden power fluctuation is detected in the power data, it is determined that the safety assessment value needs to be optimized.

[0111] When it is determined that there are no power mutation values ​​in the power data, it is determined that there is no need to optimize the safety assessment value.

[0112] Understandably, power data is a crucial indicator of battery operating status. Analyzing, denoising, and filtering this data removes interference, making it more accurate and reliable. Calculating power fluctuation indicators, such as amplitude fluctuation rate (reflecting the degree of power amplitude change over a short period), relative change rate (reflecting the proportion of power data change relative to the normal state), and duration (indicating the duration of abnormal power changes), and comparing these indicators with preset thresholds, effectively identifies whether power data exhibits abrupt changes. The presence of abrupt changes indicates a potentially significant change in battery operating status, possibly due to internal faults or external factors, meaning the original safety assessment value may not accurately reflect the actual safety situation and requires optimization. Conversely, the absence of abrupt changes indicates a relatively stable battery operating status, and the original safety assessment value is highly reliable, requiring no optimization. This method of determining whether to optimize safety assessment values ​​based on power data improves the accuracy and reliability of battery safety assessments. In practical applications, operators can accurately assess battery safety status based on the optimized safety assessment values ​​and take timely measures such as adjusting operating parameters, maintenance, or battery replacement to ensure safe and stable battery operation and reduce the occurrence of safety accidents.

[0113] Specifically, when determining the optimization coefficients for the safety assessment value based on the air pressure and electricity data inside the warehouse, and obtaining the optimized safety assessment value, the following steps are taken:

[0114] The air pressure inside the chamber was analyzed to obtain the rate of change of air pressure.

[0115] The power data is analyzed to obtain real-time current values;

[0116] The pressure change rate and pressure change threshold are compared, and the real-time current value is compared with the current threshold. The optimization coefficient is determined based on the comparison results.

[0117] When the rate of change of air pressure is greater than or equal to the air pressure change threshold, and the real-time current value is greater than or equal to the current threshold, the optimization coefficient is determined as the first optimization coefficient.

[0118] When the rate of change of air pressure is greater than or equal to the air pressure change threshold and the real-time current value is less than the current threshold, the optimization coefficient is determined as the second optimization coefficient.

[0119] When the rate of change of air pressure is less than the air pressure change threshold and the real-time current value is greater than or equal to the current threshold, the optimization coefficient is determined as the third optimization coefficient.

[0120] When the rate of change of air pressure is less than the air pressure change threshold and the real-time current value is less than the current threshold, the optimization coefficient is determined to be the fourth optimization coefficient.

[0121] The product of the optimization coefficient and the safety assessment value is used as the optimized safety assessment value.

[0122] Understandably, the optimization coefficients are ranked as follows: First optimization coefficient < Second optimization coefficient < Third optimization coefficient < Fourth optimization coefficient. When the rate of change in air pressure is greater than or equal to the air pressure change threshold, and the real-time current value is greater than or equal to the current threshold, it indicates that the air pressure inside the battery compartment is fluctuating drastically and the current is high, suggesting the battery may be in a relatively dangerous operating state. In this case, the safety assessment value needs to be significantly lowered, hence the first optimization coefficient is the smallest. When the rate of change in air pressure is greater than or equal to the air pressure change threshold, but the real-time current value is less than the current threshold, although the air pressure change is abnormal, the current is relatively small, reducing the degree of danger; therefore, the second optimization coefficient is relatively larger. When the rate of change in air pressure is less than the air pressure change threshold, but the real-time current value is greater than or equal to the current threshold, it indicates that the air pressure is relatively stable, but the current is high, still posing a certain safety risk; therefore, the third optimization coefficient is larger. When the rate of change in air pressure is less than the air pressure change threshold, and the real-time current value is less than the current threshold, it means that both the air pressure and current inside the battery compartment are in a relatively stable state, and the battery's safety condition is good; in this case, the fourth optimization coefficient is the largest. By multiplying the optimization coefficient by the safety assessment value, an optimized safety assessment value is obtained, which more accurately reflects the actual safety status of the battery. In practice, staff can use this optimized safety assessment value to manage and maintain the battery more scientifically. For example, if the optimized safety assessment value is low, a comprehensive inspection of the battery is necessary immediately to identify potential safety hazards and take appropriate measures to ensure the safe operation of the battery and prevent accidents. Conversely, if the optimized safety assessment value is high, the battery inspection cycle can be appropriately extended to improve work efficiency.

[0123] Specifically, when determining the safety level of a battery under test based on optimized safety assessment values, the following are included:

[0124] The optimized safety assessment value is compared with the preset safety level mapping table, and the safety level of the battery to be tested is determined based on the comparison results.

[0125] Understandably, a preset safety level mapping table is a pre-defined table that clarifies the correspondence between different optimized safety assessment value ranges and their corresponding safety levels. This table is based on extensive experimental data and practical application experience, and it specifies in detail the optimized safety assessment value ranges corresponding to each safety level. For example, when the optimized safety assessment value is in a relatively high range, the corresponding safety level might be "Safe," meaning the battery under test is currently in good operating condition, the probability of a safety accident is extremely low, and it can be used normally for a period of time, requiring only routine maintenance. When the optimized safety assessment value is in a medium range, the corresponding safety level might be "Caution," indicating that although the battery can still function normally, there is a certain potential risk, requiring staff to increase the frequency of battery monitoring and closely monitor changes in various battery parameters. When the optimized safety assessment value is in a low range, the corresponding safety level might be "Danger," indicating that the battery has developed a serious problem and a safety accident could occur at any time. The battery must be immediately taken out of service and thoroughly inspected and repaired to eliminate potential safety hazards. By setting up a safety level mapping table, abstract optimized safety assessment values ​​can be transformed into intuitive safety levels, making it easier for staff to quickly and accurately determine the safety status of the battery and take appropriate measures.

[0126] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0127] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A power battery safety testing and evaluation system based on multimodal visualization, characterized in that, include: The system includes a testing device and an evaluation device, with the testing device connected to the evaluation device. The testing device comprises a sealed test chamber, a safety triggering and loading module, a multimodal parameter acquisition module, and a data processing and visualization module. The evaluation device includes an analysis module, a judgment and optimization module, and a decision module. The battery to be tested is placed inside the sealed test chamber; The safety triggering and loading module is used to apply trigger loading to the battery under test under controlled conditions and obtain the response data of the battery under test under the trigger loading conditions; The multimodal parameter acquisition module is used to acquire response data and perform preprocessing. The data processing and visualization module is used to draw superimposed thermal field maps, flue gas isosurfaces, and deformation curves of the battery under test based on the preprocessed response data. The analysis module is used to extract the temperature features of the superimposed thermal field map, the flue gas features of the flue gas isosurface, and the deformation features of the deformation curve. Based on the temperature features, flue gas features, and deformation features, a coupled feature vector is constructed, and the safety assessment value of the battery to be tested is determined according to the coupled feature vector. The optimization module is used to collect the power data of the battery under test and determine whether the safety assessment value should be optimized based on the power data. If so, it collects the air pressure inside the sealed test chamber, determines the optimization coefficient of the safety assessment value based on the air pressure inside the chamber and the power data, and obtains the optimized safety assessment value. The determination module is used to determine the safety level of the battery to be tested based on the optimized safety assessment value; A coupled feature vector is constructed based on temperature characteristics, flue gas characteristics, and deformation characteristics. When determining the safety assessment value of the battery under test based on the coupled feature vector, the following is included: Temperature characteristics, flue gas characteristics, and deformation characteristics are normalized separately to obtain normalized characteristics; Feature enhancement is performed on the normalized features to obtain enhanced features; The enhanced features are aligned temporally and spatially to form an aligned multimodal feature set; Based on the multimodal feature set, construct the coupling terms between modes, and combine the coupling terms with the multimodal feature set to form the coupling feature vector; The coupling feature vector is compared with the historical coupling feature vector set, and the safety assessment value of the battery to be tested is determined based on the comparison result. The coupling terms include the coupling terms of temperature change rate and flue gas diffusion gradient, the coupling terms of flue gas concentration peak and deformation accumulation rate, and the coupling terms of temperature spatial abrupt change and deformation inflection point time sequence difference.

2. The power battery safety detection and evaluation system based on multimodal visualization according to claim 1, characterized in that, Controlled conditions include any one or more combinations of overcharging / over-discharging, heating, squeezing, and puncture of the battery under test.

3. The power battery safety detection and evaluation system based on multimodal visualization according to claim 2, characterized in that, When plotting the superimposed thermal field map, flue gas isosurface, and deformation curve of the battery under test based on the preprocessed response data, the following steps are included: The response data includes temperature data, flue gas concentration data, and deformation data; The preprocessed temperature data is processed, and thermal imaging algorithms are used to map temperature values ​​at different times and locations onto a two-dimensional plane. The temperature is represented by the intensity of color to create an overlayable thermal field map. The preprocessed flue gas concentration data is processed, and the isosurface generation algorithm is used to determine the spatial distribution boundary of the flue gas and draw the flue gas isosurface. The preprocessed deformation data is processed, and a deformation curve is plotted by fitting a curve with time as the horizontal axis and deformation as the vertical axis.

4. The power battery safety detection and evaluation system based on multimodal visualization according to claim 1, characterized in that, When determining the safety assessment value of the battery to be tested based on the comparison results, the following are included: When there is a historical coupling feature vector in the historical coupling feature vector group that is the same as the coupling feature vector, the historical security assessment value corresponding to the historical coupling feature vector is used as the security assessment value. When there is no historical coupling feature vector in the historical coupling feature vector group that is the same as the coupling feature vector, the state offset rate of the battery to be tested is calculated based on the coupling feature vector, and the safety assessment value of the battery to be tested is determined based on the state offset rate.

5. The power battery safety detection and evaluation system based on multimodal visualization according to claim 4, characterized in that, When calculating the state offset rate of the battery to be detected based on the coupling feature vector, the following is included: Obtain the time series of the coupled feature vectors, and calculate the Euclidean distance between the coupled feature vectors at adjacent time points based on the time series of the coupled feature vectors to obtain the coupling change. The coupling changes are weighted, with the weights consisting of temperature characteristic weights, flue gas characteristic weights, and deformation characteristic weights, to obtain the weighted coupling changes. The rate of change of the weighted coupling change is calculated within a preset time window to obtain the state offset rate.

6. The power battery safety detection and evaluation system based on multimodal visualization according to claim 5, characterized in that, When determining the safety assessment value of the battery under test based on the state offset rate, the following are included: The state offset rate is compared with the first state offset rate and the second state offset rate, and the safety assessment value is determined based on the comparison result; wherein the first state offset rate is less than the second state offset rate. When the state offset rate is less than or equal to the first state offset rate, the security assessment value is determined to be the first security assessment value. When the state offset rate is greater than the first state offset rate and less than or equal to the second state offset rate, the security assessment value is determined to be the second security assessment value. When the state offset rate is greater than the second state offset rate, the security assessment value is determined to be the third security assessment value.

7. The power battery safety detection and evaluation system based on multimodal visualization according to claim 6, characterized in that, When determining whether to optimize the safety assessment value based on power data, the following should be included: The power data is analyzed, and noise reduction and filtering are performed on the power data; Power mutation indicators are obtained based on filtered power data. These indicators include amplitude mutation rate, relative change rate, and duration. The power mutation index is compared with a preset threshold. Only when the power mutation index meets the preset amplitude threshold and exceeds a preset number of times within a preset time window is it determined that there is a power mutation value in the power data. When a sudden power fluctuation is detected in the power data, it is determined that the safety assessment value needs to be optimized. When it is determined that there are no power mutation values ​​in the power data, it is determined that there is no need to optimize the safety assessment value.

8. The power battery safety detection and evaluation system based on multimodal visualization according to claim 7, characterized in that, The optimization coefficients for determining the safety assessment value are based on the air pressure and power data inside the warehouse. When obtaining the optimized safety assessment value, the following steps are taken: The air pressure inside the chamber was analyzed to obtain the rate of change of air pressure. The power data is analyzed to obtain real-time current values; The pressure change rate and pressure change threshold are compared, and the real-time current value is compared with the current threshold. The optimization coefficient is determined based on the comparison results. When the rate of change of air pressure is greater than or equal to the air pressure change threshold, and the real-time current value is greater than or equal to the current threshold, the optimization coefficient is determined as the first optimization coefficient. When the rate of change of air pressure is greater than or equal to the air pressure change threshold and the real-time current value is less than the current threshold, the optimization coefficient is determined as the second optimization coefficient. When the rate of change of air pressure is less than the air pressure change threshold and the real-time current value is greater than or equal to the current threshold, the optimization coefficient is determined as the third optimization coefficient. When the rate of change of air pressure is less than the air pressure change threshold and the real-time current value is less than the current threshold, the optimization coefficient is determined to be the fourth optimization coefficient. The product of the optimization coefficient and the safety assessment value is used as the optimized safety assessment value.

9. The power battery safety detection and evaluation system based on multimodal visualization according to claim 8, characterized in that, When determining the safety level of a battery to be tested based on optimized safety assessment values, the following are included: The optimized safety assessment value is compared with the preset safety level mapping table, and the safety level of the battery to be tested is determined based on the comparison results.

Images